As you well know, I’ve been going off a lot lately on how Canonical and Ubuntu made quite the mistake with Unity. Not only has it seemingly had the opposite effect of its name, it was quite poorly designed. Unity is filled with design flaws (such as the window menu system and the horrible new scroll bars that don’t always work) that make its usage less than ideal.

But a fraction of the open source community sees that, instead of revolting, why not attempt to offer up suggestions to help improve the new Ubuntu default desktop. This “movement”, started by Arian van Gend, is called “Community Proposed Unity Design”. It doesn’t have a website, or a slogan, or even a clever graphic. Instead, what it has is a document outlining possible changes that should be made to Ubuntu Unity in order to improve it’s functionality. The goal of the document is not to attack the Unity designers, but instead to support them by offering a “centralized community design document that streamlines suggestions for users.” And, in the spirit of open source, the document is, as you would expect, open source (it’s a Google Doc that anyone can view, and either email suggestions or use the Google Chat function to chat with the author about your suggestion.)

Some of the improvements include:

Different Dash views (the Dash is the “dashboard” by which searches and applications can be found). By default the dash view is clunky and needs serious improvement. The community sees this and highly suggests different types of dash views be available. One particular Dash improvement that should seriously be considered is:

The Dash should only be accessible from the Ubuntu button on the top left. This focuses the Dash functionality in one place, and avoids cluttering the launcher, which should be used only for applications, Lens Quick Drops, and their state. A Lens Quick Drop is an Launcher item that can perform some action on objects of a certain type.
And this improvement would make using the dash much more user-centric:
Clicking the Ubuntu button should open the Dash in its universal search state. It should also change the Launcher content to a list of Lenses, displayed in a similar fashion as other items on the Launcher. The first of these Lenses should always be the Universal search Lens.
The other Lenses will be manipulable in the same way application launchers would be, meaning they can be reordered, removed, and have quick lists.
The document goes on from there, discussing the different types of lenses (various views used by the dash) such as:
  • Application lens
  • Search
  • Files/folders
  • Contacts
  • Conversations from social networking
When dealing with the Global menu, the document nails the issues:
  • Touch unfriendly.
  • Hard to discover for new users.
  • Hard to quickly find the right menu (it’s not visible initially).
  • Application title looks strange when cut off.
  • Menu seems too separated from the application sometime.
  • Root apps aren’t integrated.
  • Actually Global Menu show the application menu for the active Windows.
The possible solutions to these issues the document suggests are:
  • Show the separation between application title and menu differently;
  • Always show the menu itself (it’s not that distracting since our focus is usually somewhere else. This also fixes the findability issue and is more touch friendly;
  • Make it easy to switch application focus from the application name in the panel. This makes it easy to reach a menu on an application, different from the focused one, and mitigates the time lost when traveling up to the menu and the wrong application was focused.
  • All our windows must integrate to Global Menu, not root’s one distinction.
  • A good solution will be to use Global Menu for applications, like MacOS does. This has major benefits for applications with multiple windows, like Empathy (Contact and Chat) or the GIMP. In this case, if we are chatting and the Contact Windows are not focused,  we don’t have to make it active before accessing its menu.

The document continues on to discuss:

  • Third-party and existing indicators
  • Dynamically adjusted workspaces
  • Overlay scroll bars
  • Keyboard shortcuts/mouse gestures/touch actions
  • Pan-space (alternative to multiple workspaces)
  • Minor feature tweaks

So, yeah…this community really gets the issues surrounding Unity. But, unlike some communities, the open source community knows how to drive what they’ve been given to newer and better places. This highlights the true spirit of the open source community — not only what they are striving for, but how they are approaching their task.

I applaud Mr. van Gend and his effort. I certainly hope the Unity developers look at his document with an eye on improving the Unity desktop, as the suggestions made within are quite good.

What do you think? Do you have suggestions for the Unity desktop that should make it into Arian’s document? If so, share them with us, before you share them with van Gend.